Hi all.
Currently looking at a process where the output metric is “finish time”, which is continuous. One of the factors under consideration is a possible resend loop within the process. First cut of data is showing that c.20% of runs require a resend, and some multiple resends.
Plan to run a t-test or Kruskal Wallis to determine if there is a difference in finish time between those runs which require a resend and those that don’t.
However, I also have data on- if multiple resends were required- how many were required. Recorded figures for this have so far been 2, 3 or 4 and this is the data I need advice on. Specifically:
– Can this X of no of resends be termed continuous data? If not, what is it?
– Is a correlation regression test therefore appropriate? If not, what would be?
Many thanks,
Paul
.
.